This application is submitted in response to PA-02-077, "Secondary Analyses in Diabetes, Digestive and Kidney Diseases". We propose to analyze data from an existing observational cohort database to assess the impact of different types of contraception on the risk of diabetes and on glucose levels, blood pressure and serum lipids in Latino women with prior gestational diabetes mellitus (GDM). We have previously shown that those women are at high risk of developing type 2 diabetes during their reproductive years. They require effective contraception because a) pregnancy planning is required to minimize the risk of conceiving with undetected hyperglycemia, which imparts a risk of congenital malformations to the baby and (b) additional pregnancies increase their risk of diabetes. Contraception should be safe regarding effects on glycemia and risk of diabetes, as well as on blood pressure and lipid levels, which can be elevated in association with the insulin resistance that is common in women with GDM. Using data from a subset of the current cohort available through 1994, we found that low-dose combination oral contraceptives (OCs) did not increase the risk of diabetes, but that a progestin-only oral preparation did increase that risk. Effects on lipids and blood pressure were not evaluated. Approximately 20% of hormonal contraception users in our patients elect to use an injectable progestin preparation, depomedroxyprogesterone acetate (DMPA) for long-term contraception. The impact of that preparation on the risk of diabetes is unknown. Accordingly, we now propose to use an expanded version of the cohort, including all patients through 1998 when accrual ended, to assess the impact of DMPA as compared to OCs and non-hormonal contraception, on the risk of diabetes and on glucose levels, blood pressure and lipid levels. Survival analysis will be use to compare diabetes rates among forms of contraception. Generalized linear models will be used to compare effects on glucose, blood pressure and lipid levels. Successful completion of these analyses will provide novel and clinically important information about the safety of several forms of hormonal contraception that are crucial for young Latino women at high risk for type 2 diabetes. The observational results will also be important to the design of future direct comparisons of different forms of contraception in this high-risk population. [unreadable] [unreadable]